YT YouTube2Text

YouTube2Text → Transcripts → Greg Isenberg’s Step-By-Step Blueprint To Building A Successful AI Business

Greg Isenberg’s Step-By-Step Blueprint To Building A Successful AI Business — Transcript

by The Next Wave - AI and the Future of Technology · 7,059 words · 1,057 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00So, I'll give you the playbook of if I
0:03was starting an AI company, here's how I
0:05would do it.
0:06How do you decide what companies to
0:08either invest in or build with this
0:11whole narrative that these AI companies
0:14don't have a moat?
0:15I think it's wrong. We started a
0:16business that is going to do two to
0:18three million dollars of free cash flow
0:21without raising one cent from anyone
0:22else, and it was all using this
0:25framework.
0:27Hey, welcome to the Next Wave podcast.
0:29My name is Matt Wolf. I'm here with my
0:30co-host Nathan Lands, and we've got an
0:33excellent show for you today. So, in the
0:35AI world, you've probably heard a lot
0:37that AI businesses have no moat. It's
0:39really, really difficult to create a new
0:41business in the world of AI when
0:43everybody can create similar businesses.
0:45Well, on today's show, we've got Greg
0:47Eisenberg, and Greg has a completely
0:50different idea about how this is all
0:52going to play out. In fact, he gave us a
0:54step-by-step roadmap, a playbook on how
0:58you can build a successful business in
1:00this current world of AI. So, let's go
1:02ahead and just dive in with Greg now.
1:05Greg Eisenberg, thanks for joining us
1:07here today.
1:08It's a dream come true.
1:09This is going to be fun. I mean, right
1:10now we're in this this era where AI is
1:13sort of taking over all of the news
1:15cycles, and everything has AI baked into
1:18it. Even all the products that don't
1:19seem like they need AI have AI baked
1:22into them. And you're doing some amazing
1:24stuff with AI, so it just it seems like
1:25a good fit for you to be one of our
1:27first early guests on this show and and
1:29talk about where all of this is headed.
1:31The interesting thing about AI is
1:33everyone from your barber to Jeff Bezos
1:36is interested in AI.
1:38Yeah.
1:39I recently got a haircut, and my barber
1:41was kept asking me about different AI
1:43tools that he was using. He wanted my
1:46feedback on it. And what Nathan and I
1:48actually have in common is I think I
1:50tweeted something about AI, and he
1:52tweeted something about AI, and then
1:54within 24 hours, Jeff Bezos both
1:56followed us on Twitter.
1:57And then Matt also got followed off of
1:59that.
1:59Yeah, we're all in an exclusive Jeff
2:00Bezos follows us club now.
2:02Yeah, yeah. So we got we got to finally
2:04message him and actually get him to come
2:05on or do something with all of us, so.
2:07That's right.
2:08So Greg, how do you see the holding
2:11company concept? How do you decide what
2:13companies to either invest in or build
2:16with this whole narrative that these AI
2:19companies don't have a moat, knowing
2:21that almost any company out there can
2:24use some of these existing resources to
2:27build the same software you're building?
2:29How is the approach to what companies to
2:31invest in or build?
2:33So there's this narrative that if you're
2:35building on top of OpenAI, ChatGPT, or
2:39this wrapper, that you have no moat and
2:42therefore you haven't built any value.
2:44And I think it's wrong. I think it's
2:46wrong. Because if you think of ChatGPT
2:48and OpenAI, what is it? It's basically
2:51an amalgamation of the craziest amount
2:54of worldwide data in one place. It's
2:56it's it's incredible. It's like a
2:58trillion times the uh great library in
3:02Alexandria. Maybe even more.
3:05And
3:06I think that uh what people don't get
3:08is, you know what's another little
3:10startup that is a thin wrapper around
3:12the world's information? Uh it's Google.
3:15Google doesn't create any of the
3:18content. They are a wrapper around the
3:21internet, but they point people to pages
3:24within the internet. And it's the
3:26greatest business model of all time. You
3:29know, even today, you know, it's a
3:31trillion-dollar company, if you look at
3:33their financial statements, a huge chunk
3:35of their business still comes from
3:36Google Search. I also think that there's
3:40this great unbundling of ChatGPT that is
3:43happening. Now, what unbundling is,
3:46there's a great quote, I think Jim
3:47Barksdale said it, there's two ways to
3:49make money. You're either unbundling or
3:52you're bundling.
3:53And what that means is, what does
3:55unbundling mean? Unbundling means, you
3:58know, if you look at Craigslist,
4:00Craigslist was a marketplace for
4:02everything. You had the ability to post
4:04a job. That became indeed.com.
4:07You had the ability to, you know, find a
4:11mate. And that became, you know, Tinder
4:14and and what Match is doing. Basically,
4:17on the internet, you can't really be
4:19everything everyone.
4:21So, there's going to be a huge
4:23opportunity to look at what are the
4:25different use cases on the horizontal
4:27platforms
4:28like ChatGPT and what OpenAI is doing.
4:31And basically, think about how can I
4:34apply a community-first approach,
4:37create what you might call as a thin
4:39wrapper, but what I call is
4:42this is dope. I don't have to rebuild
4:44everything from scratch.
4:45And and just focus on a particular use
4:50case. Now, I want to give one quick
4:51example to that. I forget the name of
4:54the founder of pdf.ai. And
4:57when ChatGPT introduced this idea around
5:00you can now talk with a PDF, basically,
5:03everyone basically assumed that this
5:05guy's {quote} thin wrapper was going to
5:08go to zero. And what he noticed is that
5:10revenue actually didn't really go down.
5:13A lot more people now knew about
5:15ChatGPT. There was news about it. They
5:17knew that you can use PDFs. And he had a
5:20purpose-built
5:21business focused on it. So, I think
5:24you're going to see more of like the PDF
5:26AI stories than you are going to see
5:28other stories.
5:29Yeah, I've always thought the critique
5:31on like the thin wrapper thing was kind
5:33of Like, I don't know like it
5:35it doesn't make a lot of sense to me.
5:35Like, like almost every company on the
5:37internet's relying on something. I mean,
5:39at one level or not. I'm like, even
5:40everybody's relying on the internet.
5:41You're right. You know, where I think
5:43the community approach that you're doing
5:45makes a lot of sense It's probably a
5:47really unique advantage is that with the
5:50new AI tools and especially let's say
5:52GPT 5 comes out and it's even way more
5:54impressive than GPT 4, we're going to
5:56get to a point where you're going to be
5:58able to look at a a SaaS website and
6:00say, "Hey GPT 5, go build that for me.
6:03Make my own version for me." That might
6:05be like a year away. And if you have a
6:07brand and a community people who trust
6:09you, like you, I think they'll keep
6:11doing business with you. Whereas if
6:12you're just like a soulless company that
6:15doesn't have that, I think you're at a
6:16great disadvantage in this next wave.
6:19Uh yeah, I want to expand on that cuz I
6:21don't think a lot of people are talking
6:22about like the marketing community angle
6:25to these products. Yeah. For the longest
6:27time people were saying that energy
6:30drinks were actually a saturated market.
6:33And then out of nowhere, what's like the
6:35best performing stock of the last 12
6:36months or one of the best is Celsius
6:39Holdings. And you know, it's like a $15
6:42billion company. It's a energy drink
6:45that they're they originally targeted
6:47women. Cuz in their research and
6:49insight, they noticed that if you look
6:50at like Monster Energy and all these
6:53you know, energy drinks, it was very
6:54like male-dominated. Now, what's the
6:57difference from a formula perspective of
7:00a Monster versus like a Bang or
7:03whatever? Like probably not that
7:05different. Same with beer. You know,
7:08what's the difference between
7:09you know, a Miller Lite and a Coors
7:11Light? Now, a lot of people listening
7:14are going to be like, "No, I'm a Miller
7:16I'm a Miller I'm a Miller girl or Miller
7:18guy or I'm a Coors Light guy."
7:21But that's identity playing into it. And
7:23you you've in your identity, you're
7:25like, "I'm that type of person." And I
7:26think you're going to see that
7:27identity-based consumption model happen
7:30with AI startups, too. Like I'm not I'm
7:33a Gemini person. I'm not a ChatGPT
7:35person.
7:36Was it pdf.ai?
7:38What what would be some like advice you
7:40would give to that company as as far as
7:42like building a community around it. How
7:43do you build a community around a tool
7:45that essentially just helps you
7:47understand PDFs better?
7:49Yeah. Well, I think that the issue with
7:52pdf.ai is it's niche, not what I call
7:55super niche.
7:56So,
7:58step one is your horizontal, you have no
8:00niche, you're chat GPT. Step two is your
8:02niche, you're you've picked like PDF as
8:06your category. And then step three is
8:08micro niche, which is like your PDF for
8:10lawyers.
8:11Your PDF for accountants.
8:14What I would be doing is be focusing on
8:16the community piece if I was him. What
8:18are the different communities that need
8:21PDFs the most? And trying to retrofit it
8:24for those people cuz those people, like
8:26accountants or lawyers for example, will
8:29have a set of pain points that are going
8:31to be different than me or you just cuz
8:34we're we're different. And also they
8:36just might just trust you more if it's
8:38called legal PDF as well, which is like
8:41the the beer example. They might just
8:43trust you because they like the name,
8:45which also is an a thing that not many
8:47people talk about in startups, which is
8:49names actually pay a
8:51and you know this, Nathan.
8:54Yeah.
8:55Mr. Lawyer, but names actually pay a
8:58bigger role in how people actually
9:01connect with products than people like
9:03to think.
9:04Yeah, totally. I think you're right.
9:06Like especially like in the age of AI,
9:07that's going to matter more and more.
9:09Like having a name and a brand that
9:10people trust and they feel connected to
9:12so they keep going back and doing
9:13business with that company versus the
9:16random company that just, you know,
9:17sprouted up out of nowhere and just
9:19copied you with AI.
9:20You you bring up a good point around if
9:22if everyone could basically press the
9:24duplicate button. Yeah. Like how do we
9:26think about building companies in that
9:28in that world? It's a a really good
9:30point. I've thought about it a lot and I
9:31think that
9:32you know, the good side of it by the way
9:34is that it's going to be really easy to
9:36spin things up. Yeah.
9:38You can You can see a viral tweet and
9:39then be like, "Wow, this is really
9:40cool." And you can be living in like Sri
9:42Lanka, and you can be 14 years old, and
9:44you can press the duplicate button, and
9:46then you can like, you know, pay for X
9:49blue, and you you show up, and you
9:50reply, and you get a million downloads.
9:52Like, your life could could change
9:54overnight.
9:54Yeah. The downside of the of that coin
9:57is that competition is going to 100X.
10:01Like, the amount of software that's
10:02going to be created is going to be 10 to
10:04100X minimum.
10:06Just like you saw that in content land
10:08with tools like Tik Tok. You know, I was
10:11an advisor to Tik Tok for a few years,
10:14and it was crazy seeing the amount of
10:16content, you know,
10:18just I mean, you see it. I'm sure you've
10:20seen played with Tik Tok. Like, there's
10:21the amount of content on all these
10:23platforms, Tik Tok, YouTube,
10:25Instagram. You give people easier tools
10:28to create things, they're going to
10:29create things. So, that's what's going
10:30to happen. I'd love to talk about you
10:32guys, like, how would you navigate this
10:34world where everyone's given the
10:36duplicate button?
10:37Well, I mean, Greg, that's that's a big
10:39reason like I thought you'd be a great
10:40guest on the show, honestly, is cuz like
10:42I think your approach is right on. Like,
10:44you're building a personal brand. I
10:45think in the age of AI, that's really
10:47important. You're an actual person.
10:49People like you, and you can leverage
10:51that across various companies that you
10:52build. So, I think that that approach is
10:55right on. The community aspect, having
10:57people involved in the actual product,
10:59so they like you, and and and they feel
11:00engaged, and actually you adjust the
11:02product based on feedback and and
11:04talking with them. I think it's huge.
11:06And also, you've got the Innovation
11:07Agency as well, right? That's another
11:09one where like Yeah, in this new age,
11:11like, companies have to be thinking
11:12about, "Oh, I don't just have this one
11:14product, and people are going to use
11:15this same product for 100 years."
11:17Somebody may come and copy it, and they
11:19and they, you know, I think that I have
11:21a better brand, but maybe all of a
11:22sudden they're cool, everybody switches
11:23over,
11:25you know? And so, you have to be you
11:26have to continue innovating and trying
11:27new things. And I I think the companies
11:29that are really going to be the big
11:31companies in the next, you know, 100
11:32years are the ones that build innovation
11:34into their company and they're
11:36constantly innovating and spinning out
11:37new projects.
11:38So, I I run this site called Future
11:40Tools, which, you know, it it curates
11:42all of the latest AI tools that um that
11:44come out. There's a submission form on
11:46the site where people submit their tools
11:49to me. I review them with a team of a
11:51couple extra people, and this the tools
11:53that I think are cool make it onto the
11:55site. So, I'm truly trying to curate
11:56now. The problem that I tend to have is
11:59that
12:00every single day 13 of the exact same
12:02tool pops onto the site, right? Almost
12:04every single day I'm like, this I I I
12:06see a tool that's like, here is your AI
12:08copywriter. Use this to write your sales
12:11copy for your business. Here's an AI
12:13blog writer. Every single day there's
12:15like 11 of those submitted to the point
12:18where I can't even tell the difference,
12:20and a lot of times it's like a
12:21completely different company, but I
12:23swear I've seen that site, that UI, that
12:25that page before. And so, um
12:28you know, from the perspective I'm in,
12:30where I'm seeing all these tools being
12:31submitted to me, it feels like uh this
12:34this AI world has kind of gotten to a
12:36place where everybody is just kind of
12:38already cloning everything else. Oh,
12:40that tool worked, let's clone it.
12:41Somebody else clones it. Somebody else
12:42clones it. It gets to a point where, you
12:45know, the the the first movers, the the
12:48the ones that put out the product first
12:50are still the ones that people talk
12:51about, but then there's just like a
12:53whole bunch of junk that followed behind
12:55it. So, I do think that being an early
12:58mover is important, but then I also
13:01think about that the sort of big
13:02incumbents, right? You've got the the
13:04Googles, the Adobes, the Microsofts,
13:06companies like that. And
13:09you know, I feel like we might be moving
13:11into this world where whenever something
13:13works really, really well as a small
13:15SaaS company, one of those big companies
13:17is either just going to acquire it or
13:19build it in, and then wipe out pretty
13:22much everybody else that was doing that
13:24already. I mean, we were talking even
13:25before we hit record about uh the new
13:28open AI video model. It's like when that
13:31gets released to the public in a single
13:33day, that's already better than Runway,
13:35than Pika, than animate diffusion, all
13:38of that stuff. There's already a better
13:40option out there. So now, you know, what
13:42what do we do with those now?
13:44So, I'll give I'll give you and and
13:46folks listening the playbook of if I was
13:49starting an AI company, here's how I
13:51would do it.
13:53And I'll use an example that we've done
13:56and it has worked. Like I can finally
13:59say it's worked. We started a business,
14:01an AI-powered business that is going to
14:04do two to three million dollars of free
14:05cash flow without raising one cent from
14:08anyone else.
14:09And it was all using this framework. So,
14:12the business is called
14:13boringmarketing.com.
14:14We started off as a Twitter account,
14:17literally. Anyone could sign up for a
14:18Twitter account. It's free. More or
14:20less.
14:21And we I called it @boringmarketer.
14:25And we created a character
14:27behind this idea around, you know,
14:29people talk about boring businesses, but
14:31not that many people talk about boring
14:33marketing. So you start Step one is you
14:35you build a character. You You start
14:37with a character that people are going
14:38to connect with or maybe you are that
14:41character. Like maybe like for me like I
14:43like being that character, you know? Or
14:45or Nathan likes being that character.
14:47So, we just started like talking about
14:49boring ways to grow your internet
14:51business. People kept asking us about
14:54SEO, SEO, SEO. And then we looked at the
14:57market and we realized that there was an
15:00opportunity to create AI-assisted tools
15:04to do SEO. So we started using those AI
15:08tools on our own products. That's
15:09another benefit of having a holding
15:11company. You can basically you have this
15:12portfolio of companies that could dog
15:14food your product. The good news of it
15:17is we knew that people wanted it cuz
15:18they were telling us that they wanted
15:20it. We built this community of about
15:2310,000 people and then we took that
15:26those AI tools, we wrapped it around a
15:28service agency to to help people
15:31implement and create content and and
15:34people started seeing results. It
15:36started driving word of mouth. So now
15:38all of a sudden like fast forward like
15:40that you know, it hit a seven figure run
15:42rate within four or five months and now
15:44it's it's it's a really profitable
15:45engine and and now we're moving towards
15:48like by the time this is out like
15:50boringads.com
15:51will be out.
15:53No brand is attached to a lot of this.
15:54Like it is is the that sort of quick run
15:57rate, is it because you sort of had a a
15:59springboard with your own personal brand
16:01or do you think it's due to the the
16:03merits of the products themselves?
16:04Like I'm just going on boring marketer
16:06right now, the Twitter account, you
16:07know, 100, 200, 300 likes per tweet. And
16:11this is boring SEO stuff, right? Like
16:15What What a cool thing about having, you
16:17know, one personal brand with a few
16:19thousand people, you can kind of like
16:22give seed capital, kind of like you
16:24know, social capital to another account
16:27and be like, "Hey, like I just started
16:29boring at boring marketer, check it
16:31out."
16:31Yeah.
16:32That gets the first two or three
16:33thousand followers and then from that
16:35you try to get it to a point where it's
16:37like, you know, it starts off as a baby
16:39and then gets to a toddler and then a
16:41child and then now it's like an adult.
16:43So I think how helpful was it for me to
16:46be behind it initially? Helpful. But do
16:49I think it like you know, my take on
16:52Twitter specifically is anyone can get
16:55to 10,000 followers if you just set up
16:58tweet notifications and write thoughtful
16:59replies.
17:01So you could have done that. You could
17:02have started at boring marketer. It
17:04probably would have taken you four
17:05months to get there. I was able to do it
17:07in four days, but I don't think that
17:10should be a stopping point for people
17:11who want to do something like this.
17:13I wanted to talk about like the agency
17:15business model. I know that was one of
17:17the the notes that we had, Nathan, that
17:19we wanted to dive into. Do like agencies
17:22still exist several years down the road?
17:24I mean, if if we can if you know, chat
17:27GPT becomes our our consultant and we
17:31can prompt any website into existence in
17:34a year or two from now?
17:36I think low to mid-level agencies get
17:40commoditized to basically
17:42nothing or very little. So, you're
17:45right. In a world where you can just
17:47prompt something to create a website or
17:49a landing page or
17:51you know, a marketing video. Like of
17:53course, you know,
17:54that's tough. And I think there's
17:56millions of people who work in that
17:57space that are that's tough. Like it is
18:00a tough place to be. You know, do I
18:01think that, you know, McKinsey's going
18:04anywhere or Bain or IDEO or, you know,
18:08we have a an innovation agency where we
18:11work with Fortune 500s and and some
18:13fast-growing startups with like Jasper
18:15AI and people like that. Do I think
18:17we're going anywhere on the innovation
18:19side? Absolutely not. People forget that
18:22you need taste. Yeah.
18:24Like taste is going to be the most
18:27sought-after
18:28a skill set for the next, you know, at
18:31least 10 years, whereas, and this is
18:33somewhat controversial, I think
18:34engineering was the most sought-after
18:36skill set of the last 10 years. And I'm
18:38a I can say that as like I studied
18:40computer science. Like I'm a trained
18:42engineer, you know? But I think that if
18:44you're able to if you have good taste
18:46and you know how to prompt and you know
18:47how to come up with, you know, great
18:50ideas. I'll give you an example. Like we
18:53we came up with a really big idea for a
18:55Fortune 100 company
18:56to shift their entire business. If
18:58you're the CEO of that company, you're
19:00not just going to like prompt an idea
19:02generator to do that. You're going to
19:04want to outsource that thinking to
19:06someone that you can trust. And that's
19:07where taste comes in, right? Like
19:10So, I think that yeah, good get me taste
19:12on
19:12Yeah.
19:12You got to
19:13Another sort of example along the same
19:15lines is right I've got a YouTube
19:17channel.
19:18All of my thumbnails for my YouTube
19:20channel are created with AI. I've got
19:21like an AI model with my face trained
19:23into it. You know, I I use midjourney
19:25for backgrounds. I actually still have a
19:28thumbnail designer on my team. He uses
19:31AI for me because he knows what
19:33thumbnails look better than I do, right?
19:36So he can actually use these tools
19:38Dolly, the stable diffusion, midjourney.
19:41And I mean I can create amazing images
19:43with those. He creates really amazing
19:45thumbnails, pulls them all together, and
19:47still has a much better design eye than
19:49I have. So I still hire somebody to use
19:52AI for me to make the thumbnails.
19:54Yeah. Yeah, but I think one thing there
19:56though is like, you know, a lot of grunt
19:58work I think will be kind of replaced
20:00with this, all right? It's like your the
20:02guy who's making the thumbnails, he has
20:03great taste. But if it was somebody who
20:04was doing something just very repetitive
20:07without a lot of thought put into it or
20:08you know, required taste, a lot of that
20:11is going to get automated away.
20:12Yeah, and I think in a lot of ways too,
20:14it you know, what you were saying it
20:15sort of levels the playing field as
20:17well. So yes, there's going to be like
20:20non-skilled people that might struggle a
20:22little bit, but those people in that
20:23sort of middle-class realm, they also
20:26have these tools at their disposal to
20:29sort of take them to the next level. If
20:31you want to learn how to code, it's
20:32easier than ever right now to learn how
20:34to code. If you want to learn how to be
20:35really good at graphic design, it's
20:37easier than ever to learn how to be
20:39really good at graphic design. So I feel
20:41like the people that are sort of in that
20:43middle lower end, I know this is like a
20:45buzzword, but it sort of democratized
20:47the information to to get you there,
20:49right? Like now it's a lot easier to go
20:51from zero to one with the tools that are
20:54available out there.
20:55I think those people are at a
20:57crossroads. Like they can either see the
20:59knowledge and see the tools and be like,
21:01I'll get to you later, or they can get
21:04their hands dirty. Mhm.
21:06kind of convinced this stuff's going to
21:07get you know, 20 to 50% better every
21:10year for the next 5 years. So, that's
21:12that's coming soon that this is going to
21:13be able to help you plan out
21:15your life in like an action plan of
21:17things you could be doing to make your
21:18life better. So, I I think that's going
21:20to be such an unlock for so many people.
21:23There there's this sort of narrative
21:24going on. We were talking about the, you
21:25know, AI video generators. Like OpenAI
21:28just came out with theirs. And there's
21:29this narrative that I just absolutely
21:31hate that I see all over Twitter that's
21:33like, "Oh, it's the end for directors.
21:35Oh, it's the end for filmmakers. Look
21:37what you can do now." And I I can't
21:39stand that narrative. I don't like the
21:40narrative of anytime somebody goes on
21:42Twitter and makes a blanket statement
21:44of, "This AI is killing this industry."
21:46What I feel right now is that it levels
21:50the playing field, but it also sort of
21:53raises the playing field. It makes
21:54everybody that is
21:56sort of making videos, making art sort
21:59of up-level their game because now what
22:02used to be something people thought was
22:03really good, now anybody can do it. So,
22:06the people that were good at that now
22:08need to up their game and get even
22:10better at that thing because pretty much
22:12everybody can do the original thing. And
22:14so, I feel like that is kind of what
22:17we're what we're seeing right now,
22:19especially with like AI video and AI
22:21image creation, is just this this
22:24leveling the playing field for what used
22:26to be considered good, but also raising
22:28the bar for what is really amazing now.
22:30And I feel like it gives creatives, like
22:33the the video creators, the art
22:35creators, superpowers, right? Like if
22:37you're really, really good at art and
22:39now you have AI at your disposal to make
22:41art even better, you're still going to
22:44be ahead of the game of the person who
22:45just learned AI art.
22:47So, yeah, let's let's talk about that.
22:49So,
22:50Sora, which is the model I think you're
22:53you're talking about, is really
22:54interesting cuz you can essentially
22:56prompt it to create, let's just say I
22:58wanted to create an animation film. And
23:02I'm right now I think it's only up to 1
23:03minute 1 minute of film, but I'm I'm
23:05sure in the future you'll be get access
23:08to 90 minutes.
23:09And I wanted to create a film that
23:11essentially was like a Disney a Disney
23:14film
23:15but with my own script. Like that is in
23:17the realm of possibility in the future.
23:19So if anyone could be a Walt Disney
23:22basically
23:23and you you you can make that assumption
23:26if you have good ideas, right? If going
23:27back to the taste thing, you know you
23:29have good taste and you can write well
23:31and you have the skills to do that
23:33then what is mispriced? The mispriced
23:35piece is the distribution.
23:38The only place that you can't compete
23:40with Disney is in distribution.
23:42They're in tens of thousands of movie
23:44theaters or hundreds of thousands of
23:46movie theaters across the world. They
23:47are on Netflix. Like they've done these
23:49deals. They're taking out Super Bowl
23:51ads, stuff like that, right? So that's
23:54where I think that people are mispricing
23:56creators and they're mispricing
23:59distribution in general. So if I'm
24:03listening to this and if I'm me too like
24:05and and I'm you like the thing to do is
24:08two things. One is you build a
24:11distribution lane.
24:13So you do everything you can to build as
24:15much distribution and credibility and
24:18trust as possible.
24:20And then the other lane
24:22is playing is the best way to describe
24:26it. Like you play with the tools and you
24:28play to learn. But you have to pick you
24:31can't just be like I'm going to be the
24:33video person and the audio person and
24:35then the writer person and then you have
24:37to like pick a lane also, right? Like
24:39you have to decide like oh my dream is
24:42to build
24:43is to be the next Walt Disney example.
24:45I spent a little bit of time in
24:46Hollywood. So for like a about a year
24:48and a half I was partner with Barry
24:49Osborne. I think Greg maybe I told you a
24:51little bit about this, but we you know
24:52lord.com originally the reason I bought
24:54the fancy domain
24:56is cuz I was you know I was partner with
24:57the producer of Lord of the Rings. I was
24:58like, we're going to make this new kind
25:00of movie studio together and it was this
25:01crazy dream of mine. And so I got to
25:03spend time on the the set of Mulan out
25:06in New Zealand, you know, Disney set and
25:08got to meet all these crazy people out
25:10in Hollywood, New Zealand. And but one
25:12thing you realize is it's so hard to
25:13break into that industry, right? And
25:15like you really have to know people and
25:17things like that and it does feel like
25:18with this new technology like so many
25:20more people are going to get discovered
25:22cuz they can they can show they can show
25:24their concepts to Netflix or whoever,
25:26right? Like right now the people who
25:27pitch Netflix, they're coming in with
25:29storyboards and things like that and a
25:31team and and that's a big part of what
25:33sales sales it is like the storyboard
25:35and like the concept and the research
25:37around it and before people you had to
25:40have a whole team to do that and now
25:41like somebody who has good taste they
25:43can have those ideas, produce the
25:45storyboards or a short video concept.
25:48Maybe not maybe that's not that's not
25:49going to be the final film, but they can
25:51produce them 1 minute or 90 minute video
25:54and say, "Hey, here's not the final
25:55film, but here you get the the gist of
25:57it. Here's the gist of this idea I have.
26:00Help me make it." And I think a lot of
26:01people are like Netflix will write
26:03checks like large checks to come in and
26:05produce those films. And so that's going
26:07to be exciting for so many people.
26:08So you think that like AI video is the
26:10new storyboard instead of making a
26:12storyboard? Here's just a mock of the
26:14video.
26:14I think there's so many opportunities
26:15around that. I think I think you're
26:17you're going to see like talent agencies
26:18too where they'll like, you know,
26:20they'll they'll pitch the studios with
26:21their talent and like just put our
26:23talent's face into the concept like
26:25here's how this you know, this guy that
26:26we have, girl, this is how they would
26:28look in this role and then you just show
26:29it, you know. There's like so many areas
26:31entertainment that's going to get
26:32changed from this, yeah.
26:33I also think that the leverage that
26:35you're going to have with the Netflixes
26:37of the world is going to be oh, I've got
26:4050,000 or 100,000 followers and I've
26:44already
26:45posted this IG clip or whatever and I
26:49got 20,000 shares and they're loving it
26:52and they're they're banging on the door,
26:53and they want more of it.
26:54Yeah.
26:55So, I think that's why I always go back
26:56to the distribution piece, cuz like I
26:58think that that if everyone has the
27:00ability to create essentially these
27:02Disney-style like levels of films and
27:05and and storyboards, like it's going to
27:07be you have to have some point of
27:09leverage.
27:10Yeah, I mean, when you look at like the
27:11the music industry and even like the the
27:13book industry, right? Among authors,
27:15when you go to get like a publishing
27:16deal for books, or if you go to get like
27:19a you know, a music production deal from
27:21these these music studios, that's kind
27:24of the thing they're looking at these
27:25days is how big of a following do you
27:26have, right? It's a heck of a lot easier
27:29to get a book publishing deal if you
27:30have a million subscribers on YouTube
27:32and, you know, a 1.5 million followers
27:35on X. You know, it sounds like that
27:38might be the future of video as well. If
27:40the sort of creation of video is
27:42completely democratized and anybody can
27:44make them, that's sort of the the next
27:46level is what sort of distribution do
27:47you bring to the table on your existing
27:49platforms? That's a I think that's a
27:52really interesting way to to look at it
27:53if I'm interpreting what you're saying,
27:54right?
27:55Yes. I mean, you saw that a little bit
27:56with music, you know, 10, 15, 20 years
27:59ago even. When YouTube came out, like
28:01Justin Bieber in was a YouTube creator.
28:04You know, we didn't call him that, but
28:05he posted videos on YouTube. He was
28:07found by Scooter Braun on YouTube, and
28:10then he became
28:12Justin Bieber, one of the biggest acts
28:15in the world, musical acts in the world,
28:16from YouTube. We don't have that really
28:19with
28:20movies, for example. Like the way you
28:23become big in movies is you go on a more
28:25traditional path.
28:27Like you go to USC and study film, and
28:30then you get an internship at, you know,
28:33you know, XYZ, you know, I don't even
28:35know, cuz I don't know the film
28:36business, but I know all I know from
28:38what I hear is that it's it's a very
28:40linear path, and you get like here's
28:41like you have to like check the boxes.
28:44What this does is there's only two boxes
28:47to check. Do you know the tools and do
28:49you have distribution is what I'm
28:50saying.
28:51And you can use the tools to get
28:52distribution too is the beautiful thing,
28:54right? Like you could be somebody in
28:56Quebec you whatever like in a small town
28:58who starts producing films or or AI art
29:01or whatever on Twitter and get an
29:02audience and then go out and try to take
29:05the next level and and actually, you
29:06know, partner with some big company to
29:07actually make it.
29:08Absolutely. I also think this is kind of
29:11a
29:12little bit of a different topic, but I
29:14think there's a huge opportunity to
29:16create content for different geographies
29:19Mhm.
29:20and languages and cultures.
29:22Yes.
29:23I'm interested to see how AI plays a
29:24role in that.
29:26So that like you're not just lip dubbing
29:28different videos for example. Like what
29:30if in France you're wearing like a beret
29:32or something and you know, in
29:36and and then the video in the US maybe
29:38you're not wearing a beret, right?
29:39You're you're wearing a baseball cap or
29:40something like that. Just like what are
29:42these little nuances things that you can
29:44add to the content
29:46so that it feels more oh, this is for
29:49me.
29:50This is for people like me and I think
29:53that's really interesting and I think
29:54that the greater trend to that is
29:57all media will be personalized.
29:59I think that's the really cool part
30:01about where we are in like the cycle in
30:03the this AI world. It's like
30:06it's not just an AI world. It's like
30:08there's a lot of mac massive trends
30:11happening right now like AI uh
30:14virtual reality
30:15Yep.
30:16Robotics is just starting as well and
30:17yeah.
30:18Robotics is just starting. I saw
30:20YC came out with their like here's the
30:23things that we're really interested I
30:24don't know if you saw that.
30:25Yeah, I saw it yesterday.
30:26Yeah, it's like you know, applying
30:28machine learning to robotics using
30:29machine learning to stimulate the
30:30physical world, new defense technology,
30:33bring manufacturing back to America, new
30:35space companies, climate tech
30:37Yeah.
30:37commercial open source companies,
30:39spatial computing, new enterprise
30:41resource planning software, developer
30:43tools inspired by existing internal
30:45tools, explainable AI, LLMs for manual
30:49back office processes in legacy
30:51enterprises, AI to build enterprise
30:53software, stablecoin finance, a way to
30:56end cancer, foundation models for
30:58biological systems, the managed service
31:00organization model for health care,
31:02eliminating middlemen in health care,
31:04better enterprise grew, and finally
31:07small fine-tuned models as an
31:09alternative to giant generic ones.
31:11There's a lot of stuff going on.
31:12And and and people were saying that part
31:14of that whole list was oh, well, you
31:16know, it's the end of the trend with SAS
31:18or something like this. I
31:20I think YC's like seeing the writing on
31:21the wall that like dramatic changes are
31:24coming. Like dramatic changes. And so
31:26yeah, we should be looking for more
31:27moonshots, big things, and and and it
31:29now the technology is making it
31:30possible. But also from a, you know,
31:32probably from their perspective from a
31:34in in investment stand standing I like
31:37the companies that they would have
31:39invested in before, a lot of those now
31:41probably be built with like one or two
31:42people. And they're not going to need as
31:44much capital, you know, and and Greg you
31:45talk a lot about this kind of stuff, but
31:47so I think they're looking like, okay,
31:48yeah, you
31:49investors and VCs are probably going to
31:51focus more and more on the companies
31:52that are really going to need a lot of
31:53caps.
31:54Right.
31:54People that are trying to solve cancer,
31:55people are trying to build robotics, and
31:57things like this, and like new type, you
31:59know, simulating the world with AI, like
32:01like very ambitious things that are
32:02going to require lots of capital because
32:05the other ones just like, God, it's not
32:06going to be a year or two. And
32:08especially, you know, I've been hearing
32:10from friends in San Francisco and a lot
32:11of them are like connected to the whole
32:12YC kind of network. Really positive
32:15things about GPT-5.
32:17I assumed, you know, since, you know,
32:19Sam Altman's connection to YC, a lot of
32:21those people know
32:22ideas of what's coming. And so they're
32:24probably realizing like, yeah, in a year
32:26or two from now, a lot of those simple
32:28SAS tools, if you don't have great
32:29distribution or a a brand, they're not
32:32going to make a whole lot of sense. And
32:33so I think that's part of the list, too.
32:34Is is that
32:35Like personally, I'm happy that people
32:37are working on those hard problems, but
32:39I personally prefer working on easy
32:42problems.
32:42The one to n or or whatever versus zero
32:44to one?
32:45Yeah.
32:45Yeah, I think it's like I'm a thin
32:48wrapper guy.
32:50And I've I've I'm I did the whole
32:52Silicon Valley thing. I lived 9 years
32:54there. I've been a part of companies
32:56that have raised billions of dollars.
32:57Like I could say I did it. But man, am I
33:00glad to be on the other side.
33:02Yeah.
33:03And yeah. And and and creating thin
33:06wrappers
33:07with deep communities. Like that should
33:09be the title of this this Something like
33:12that. It's like thin wrappers and deep
33:14communities. Like that's the strategy
33:15for You want to win in AI? Like that's
33:17literally the strategy.
33:18But before we do wrap up, I do want to
33:20ask about you probably need a robot.
33:22Could you quickly explain that to me
33:24because I was asking Nathan about it and
33:27you know, I figured the the best person
33:29to ask about it and get the explanation
33:31from is from Greg himself. So could you
33:33tell me a little bit about what that is?
33:34I start all businesses with this
33:36framework of mine. We call it the ACP
33:38funnel. Audience, community, product.
33:41So like I was talking about with boring
33:42marketing, we started with the Twitter
33:43account. We did the same thing with you
33:46probably need a robot. This is like over
33:48a year ago. Yeah, over a year ago.
33:51AI obviously wasn't as big as what it is
33:53today.
33:54And we created a Twitter account where
33:57we just share productivity tips with
34:00like we use this tool and here's what we
34:02learned. And then we opened up a
34:03Discord.
34:05And then we got like 20,000 people to
34:07join the Discord. Uh maybe more, 25,000.
34:11And we saw that there was demand. And
34:13then we
34:15of those people, we turned it into a
34:16newsletter. And then every week we would
34:18give people like deep dives around
34:20here's how we're using AI this week.
34:23So less of not like AI news or anything
34:25like that. But just like here's here's
34:28things that we're learning and
34:30to make it more productive in your
34:32business career. And it's it's evolved
34:35other things, too. Like we we created
34:37like a deal pass, which is very popular,
34:39where you get access like
34:4248
34:44bucks, and like you get, you know,
34:46discounts on like all the major AI
34:48tools, or a lot of them. Like My Mind,
34:51And Drip AI, Framer, and
34:54Fairly.ai, Apollo. Mhm. So, a lot of
34:57things like that. And so, we've been
34:59able to monetize really well through
35:00this like audience community product.
35:02Mhm.
35:03And then we also, you know, cuz we run a
35:05lot of agencies, we also do helping
35:07companies like transform their
35:08businesses to be AI first.
35:10The the reason I wanted to sh- Yeah, I
35:11just wanted to share that structure, cuz
35:13I think that that same structure, ACP,
35:15could be helpful for other people.
35:17Nathan, any anything else you want to
35:18you want to ask or add before we we wrap
35:20this up?
35:21Yeah, I I really appreciate you coming
35:22here, and it's it's been awesome.
35:23Thank you for your time, and I'm excited
35:25to see where this goes, and you you
35:27you're doing great.
35:28Appreciate it.

About this transcript

This page contains the full transcript of Greg Isenberg’s Step-By-Step Blueprint To Building A Successful AI Business by The Next Wave - AI and the Future of Technology, generated from the public captions YouTube serves with the video. The transcript has 7,059 words across 1,057 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.